Smart Weed Management (SWEEM): Data-Driven Weed Emergence Modeling for Sustainable Crop Production

Overview

SWEEM (Smart Weed Management) is a data-driven approach to weed control that replaces blanket herbicide applications with targeted strategies informed by weed emergence patterns. By understanding when and how specific weed species emerge based on soil temperature and humidity (hydrothermal time), growers can time interventions more precisely—combining chemical and physical methods to maintain crop yields while significantly reducing environmental impact. The system quantifies the water and carbon footprint of each management strategy, enabling sustainability metrics to be marketed alongside the final food product. A smartphone app and supporting website deliver real-time, location-specific recommendations to growers.

Technical specifications

• Weed emergence modeling: Hydrothermal time-based emergence curves calibrated for local soil temperature and humidity conditions across multiple crop systems including fruit, vegetables, and forestry.
• Integrated management strategies: Combines chemical and physical weed control options, validated against yield outcomes to ensure crop production is maintained at lower herbicide input levels.
• Environmental footprint quantification: Water and carbon footprint metrics calculated for each weed management strategy, supporting sustainability claims for end food products.
• Decision-support app: Smartphone application linked to a web platform for continuous updates, providing growers with field-specific management recommendations.
• Broad applicability: Models developed for Chilean conditions with successful validation in additional countries; designed for scalable deployment across Latin America and other regions.

Technology readiness level

SWEEM has reached an advanced stage of development. Weed emergence patterns have already been established for local conditions in several crops in Chile, with additional successful results obtained in other countries. Research has been conducted across multiple crop systems—fruit, vegetables, and forestry—to validate management options in each context. The next phase involves selecting the most important weed species in Chile's most relevant crop systems, generating emergence curves over two full growing seasons, contrasting management strategies with yield data, integrating carbon footprint information, and deploying the final adjusted models through a smartphone app linked to a continuously updated website. The technology is positioned for pilot deployment and regional scaling.